Macromolecular X-ray Crystallography. Amie Boal Northwestern University Rosenzweig Lab

Transcription

1 Macromolecular X-ray Crystallography Amie Boal Northwestern University Rosenzweig Lab

2 References Rhodes, G. Crystallography Made Crystal Clear. 3rd ed Academic Press, Burlington. Drenth, J. Principles of Protein X-ray Crystallography Springer-Verlag, New York. Hampton Research Catalog, volume 18 McPherson, A. Crystallization of Biological Macromolecules CSHL Press, Cold Spring Harbor.

3 The real experts... Amy Rosenzweig Tom Lawton

4 Why X-ray Crystallography? Widely-used method for molecular-level 3-D structure determination of biomolecules (proteins, nucleic acids)

5 Why X-ray Crystallography? For bioinorganic chemists: Visualize how biomolecules interact with metals Structures of complex metallocofactors Identification of metal centers Stoichiometry Sommerhalter et al. (2005) Inorg. Chem. 44,

6 Why X-ray Crystallography? Limitations Oxidation states often unknown Can be affected by X-ray exposure Occupancy of metal binding sites can be altered An ensemble technique Sommerhalter et al. (2005) Inorg. Chem. 44,

7 A crystallographic experiment 2006 Academic Press 2006 Academic Press Protein Crystals X-ray diffraction Electron density map and Molecular model Why use this approach?

8 Microscopy analogy 2006 Academic Press

9 Diffraction Bending or scattering of waves around obstacles Object and wavelength (λ) must be of similar size Scattered waves can interfere constructively or destructively double slit experiment

10 Diffraction and structure In microscopy, light is diffracted by objects and is focused by lenses to create a magnified image λ of incident light ~ object Visible light is nm Bonded atoms are 0.1 nm (~1 Å) Appropriate λ is in X-ray range ( Å)

11 Two problems: 1. Single molecules diffract X-rays weakly 2006 Academic Press 1I. X-rays can t be focused by lenses 2006 Academic Press

12 A crystallographic experiment 2006 Academic Press 2006 Academic Press Crystallization Diffraction data collection Structure solution Model building and refinement

13 An X-ray diffraction experiment APS Argonne National Lab Rigaku in-house generator

14 I. Crystallization A macromolecular crystal is an ordered 3D array of molecules held together by weak interactions Made up of unit cells Cells are the repeating unit 2006 Academic Pr

15 I. Crystallization Lattice is array of unit cells Crystallographic experiment provides an image of average electron density in a unit

16 I. Growing crystals supersaturation nucleation and growth Theory: supersaturation, nucleation, growth Practice: slow controlled precipitation from solution without unfolding the protein

17 I. Growing crystals Supersaturation can be achieved by addition of precipitating agents Salts (ammonium sulfate, NaCl) Polymers (polyethylene glycol) Organic solvents (alcohols, 2-methyl-2,4- pentanediol)

18 I. Growing crystals Nucleation and growth An aggregate of sufficient size and order initiates crystal growth Usually a random event but techniques exist to drive nucleation (i.e., seeding) Additives can influence growth and nucleation

19 I. Growing crystals Other variables Protein purification, preparation, stability Structural homogeneity Temperature ph

20 I. Improved purification N-terminal His-tagged B. subtilis NrdF old prep 20-30% PEG M lithium sulfate 0.1 M HEPES ph N-terminal His-tagged B. subtilis NrdF new prep + MonoQ purification step same crystallization conditions

21 I. Growing crystals Other variables Protein purification, preparation, stability Structural homogeneity Temperature of metal cofactors ph

22 I. Growing crystals Other variables Protein purification, preparation, stability Structural homogeneity Temperature ph M. capsulatus (Bath) pmmo

23 I. Growing crystals can t predict crystallization conditions extensive screening by trial and error required

24 I. Growing crystals low [ppt] high [ppt]

25 I. Growing crystals Sparse matrix screens are typical first step Diverse sets of conditions based on database mining of published crystallization conditions Intentionally biased screen towards conditions that have worked previously Commercially available, 96-well format

26 I. Once you have crystals... Are they protein crystals? Are they crystals of the protein of interest? Is it a single crystal of sufficient size? Can the crystals be harvested and frozen? Do the crystals diffract X-rays? Can enough data be collected to solve the structure?

27 I. Optimization Secondary screens: 24-well format Homemade screens ph vs. [ppt] is a typical starting point Other variables come into play

28 200 Academic I. Crystal lattices A crystal lattice is made up of unit cells Unit cells have a defined shape x, y, z coordinate system

29 I. Lattice types 14 different lattice systems Defined by unit cell shape a = b = c relationships are identities If a = b, unit cell contents along axes are identical Can have internal symmetry - multiple copies of protein in unit cell - higher order oligomerization state related by symmetry Final model only describes structure of asymmetric unit

30 I. Symmetry operations Symmetry operations classify lattices into space groups 230 different space groups Chiral molecules restricted to 65 possible space groups Assigning the correct space group is really important!

31 A crystallographic experiment 2006 Academic Press 2006 Academic Press Crystallization Diffraction data collection Structure solution Model building and refinement

32 II. A diffraction experiment Harvest and freeze crystals Take to X-ray source Screen for diffraction Collect datasets

33 II. A diffraction experiment

34 II. Diffraction data Each spot is a reflection with an index Indices tell us lattice type and unit cell dimensions (h,k,l) index, indices increase intensity (0,0,0) 2006 Academic Press

35 II. X-ray diffraction In single crystal diffraction, only diffracted X-rays with strong constructive interference are observed

36 Acade II. X-rays are waves Have an amplitude, phase, frequency f(x) = F cos 2π(hx+α) Constructive interference increases amplitude

37 II. Bragg s law Parallel planes of atoms in crystals produce diffracted X-rays with strong constructive interference λ Conditions that satisfy Bragg s law lead to diffraction 2dhkl sin θ = nλ in phase

38 II. Planes?? A crystal lattice can be divided into regularly spaced sets of parallel planes Indexed based on number of times planes intersect the unit cell on each a,b,c edge Index notation is h, k, l and is the same as the index of corresponding reflection in the diffraction pattern

39 II. Examples of lattice planes 2006 Academic Press 2006 Academic Press

40 II. Planes II Each set of planes gives rise to a single reflection Reflections form a reciprocal lattice The reciprocal lattice is what we see in diffraction data - but related to crystal lattice via indices of lattice planes.

41 II. Real vs. reciprocal space In crystals... a = 1/a* b = 1/b* c = 1/c*

42 II. Diffraction patterns Reciprocal lattice spacing is related real lattice spacing Can calculate unit cell dimensions from h,k,l indices alone Position of spots in diffraction pattern contains information about unit cell and its symmetry

43 II. What about intensities? Intensity reports on amount of electron density associated with given set of planes Electron density associated with every atom in the unit cell contributes to the intensity of a single reflection

44 II. Diffraction data What do we look for in a diffraction pattern? resolution limit spot shape single lattice (h,k,l) index, intensity 2006 Academic Press

45 Multiple vs. single lattice edge 1.7 Å

46 II. Data collection What is a dataset? Set of diffraction pattern images collected while rotating the crystal Goal is to collect a complete and redundant set of reflections and their intensities

47 Crystal rotation during data collection captures unrecorded reflections

48 II. Data collection variables X-ray exposure time Avoid overexposure/saturation Crystal-to-detector distance Use resolution limit as guide Frame width (ω) Partial vs. full reflections and overlaps How much data to collect?

49 II. How much data to collect? # of unique reflections depends on internal symmetry of lattice Higher symmetry = less data 180 is a complete set Friedel s law: Ihkl = Ih k l exceptions...

50 II. Data processing Peak search Indexing Integration Scaling Final output is a list of reflections (h, k, l), and intensities (I) Postrefinement

51 III. Structure solution Structure factors Each reflection can be described by a structure factor (Fhkl) function Fhkl is the sum of diffractive contributions of all atoms in the unit cell no. of atoms in unit cell phase of diffracted ray atomic scattering factor

52 III. Structure solution Structure factors can be expressed as sum of contributions from volume elements in crystal electron density of volume element centered on (x, y, z) = ρ(x, y, z) electron density function volume of unit cell

53 III. Structure solution Electron density map is 3D plot of ρ(x, y, z) Fourier sum electron density function structure factor sum over all reflections in diffraction pattern

54 III. Structure solution How do we compute ρ(x, y, z) from diffraction data? Fhkl is a periodic function and has an amplitude, frequency, and phase amplitude ~ frequency = phase =??

55 III. Solving the phase problem Three common methods Molecular replacement Isomorphous replacement Anomalous scattering

56 III. Solving the phase problem Molecular replacement Use similar protein of known structure to compute phases Structure can be solved from single, native dataset

57 III. Solving the phase problem Isomorphous replacement and anomalous scattering Make use of heavy atom sites (i.e., metals) in crystals (derivative or native) Typically requires collection of multiple datasets

58 III. Anomalous Scattering X-ray absorption by heavy atoms alters diffraction Friedel s law does not hold Must be able to tune λ to absorption edge Intensity difference in Friedel pairs locates heavy atom sites in unit cell - used to compute phases Additional datasets are collected near heavy atom absorption edge Requires tunable X-ray source and data must be highly redundant with minimal X-ray damage

59 III. Anomalous scattering Can also be used to identify metals in crystal structures

60 IV. Model generation 3D plot of ρ(x, y, z) function produces electron density map Phase improvement Map interpretation Refinement of coordinates phases amplitudes (from intensities)

61 Map Interpretation Blue = 2Fo-Fc map Red/Green = Fo-Fc map 1.8 Å resolution map

62 Map Interpretation Blue = 2Fo-Fc map Red/Green = Fo-Fc map 1.8 Å resolution map

63 IV. Map interpretation 3D plot of ρ(x, y, z) function produces electron density map Phase improvement Map interpretation Refinement of coordinates phases amplitudes (from intensities)

64 IV. Model evaluation Data collection statistics Resolution Rsym/Rmerge < 0.1 I/σI > 2 Completeness/redundancy ~ 100%

65 IV. Data collection statistics

66 IV. Model evaluation Model refinement statistics R-factor Rfree statistic RMS deviation (bond lengths, angles) Ramachandran statistics

67 IV. Data collection statistics

68 IV. Model evaluation Considerations for structures with transition metals Metal ion occupancy Oxidation state and influence by X-rays Ligand order and occupancy Anomalous scattering data Evaluate by inspecting electron density maps

69 Three case studies... Crystallographic site assignment in the Mn/Fe cofactor of class Ic RNR Hogbom, M., Stenmark, P., Voevodskaya, N., McClarty, G., Graslund, A., and Nordlund, P. (2004) Science 305, Andersson, C.S., Ohrstrom, M., Popovic-Bijelic, A., Graslund, A., Stenmark, P., and Hogbom, M. (2012) J. Am. Chem. Soc. 134, Dassama, L.M.K., Boal, A.K., Krebs, C., Rosenzweig, A.C., and Bollinger, J.M., Jr. (2012) J. Am. Chem. Soc. 134,

70 Class I `2 structures 1 Class Ia 2 Class Ib Class Ic

71 Three case studies... Crystallographic site assignment in the Mn/Fe cofactor of class Ic RNR Hogbom, M., Stenmark, P., Voevodskaya, N., McClarty, G., Graslund, A., and Nordlund, P. (2004) Science 305, Andersson, C.S., Ohrstrom, M., Popovic-Bijelic, A., Graslund, A., Stenmark, P., and Hogbom, M. (2012) J. Am. Chem. Soc. 134, Dassama, L.M.K., Boal, A.K., Krebs, C., Rosenzweig, A.C., and Bollinger, J.M., Jr. (2012) J. Am. Chem. Soc. 134,

72 Mn/Fe anomalous scattering Mn edge Fe edge Andersson, C.S., Öhrström, M., Popovic-Bijelic, A., Gräslund, A., Stenmark, P., and Högbom, M. (2012) J. Am. Chem. Soc. 134,

73 Initial crystal conditions 1.92 Å Andersson et al. (2012) J. Am. Chem. Soc. 134,

74 New crystal conditions Mn edge 1.85 Å Fe edge 1.7 Å site 1 binds Mn, Fe (or Pb) site 2 binds Fe Andersson et al. (2012) J. Am. Chem. Soc. 134,

75 Site occupancy and loading A E89 1 E120 2 E193 B E89 E E193 H123 E227 H123 E227 H230 H230 Apo protein loaded with 1.5 eq Mn II and < 0.5 eq Fe II + O2 <1 eq Mn II and 1.5 eq Fe II + O2 Dassama, et al. (2012) J. Am. Chem. Soc. 134,

76 Three case studies... Structure of FeMoCo in nitrogenase Kim, J., and Rees, D.C. (1992) Science 257, Einsle, O., Tezcan, F.A., Andrade, S.L., Schmid, B., Yoshida, M., Howard, J.B., and Rees, D.C. (2002) Science 297, Spatzal, T., Aksoyoglu, M., Zhang, L., Andrade, S.L., Schleicher, E., Weber, S., Rees, D.C., and Einsle, O. (2011) Science 334, 940.

77 2.75 Å resolution Kim, J., and Rees, D.C. (1992) Science 257,

78 1.16 Å resolution Einsle, O., Tezcan, F.A., Andrade, S.L., Schmid, B., Yoshida, M., Howard, J.B., and Rees, D.C. (2002) Science 297,

79 Three case studies... Identification of the metal sites in pmmo Lieberman, R.L., and Rosenzweig, A.C. (2005) Nature 434, Hakemian, A.S., Kondapalli, K.C., Telser, J., Hoffman, B.M., Stemmler, T.L., and Rosenzweig, A.C. (2008) Biochemistry 47, Smith, S.M., Rawat, S., Telser, J., Hoffman, B.M., Stemmler, T.L., and Rosenzweig, A.C. (2011) Biochemistry 50,

80 pmmo X-ray structures 1 Cu 2 Cu XAS 1 Zn? M. caps sp. M 0.2 M ZnAc CH Cu? O2 1 Cu CH3OH OB3b Tom Lawton

81 Acknowlegements Amy Rosenzweig Tom Lawton Jason Xu Elle Moore Paula Ham Gaby Budzon Sara Sperling Kate Kurgan Tianlin Sun